MIT MechE Symposium: Mechanical Engineering and the Information Age - David R. Wallace, Anthony Patera, Seth Lloyd & Panel

[MUSIC PLAYING]

MODERATOR: As Alex d'Arbeloff mentioned in his lunch time talk. One of the main issues that we're looking at here is, okay, in some sense, it's too late for us to do anything about the fact that information technology is important for mechanical systems, right? That's going to happen whatever we do. Nobody's going to take the microprocessors out of cars because we asked them to.

But we're the people here who are educating students going to live in this world. And so what are we going to do about it? I mean, pretty clearly, the kinds of things that we've been doing up until now, though of course, they're very good, don't necessarily address these issues. So how are we going to address these issues? That's one of the questions. How are we going to include information technology in education for engineers?

And secondly, because we always do it both ways, what does information technology have to do for education? How can we use information technology in order to make education more efficient for engineers? All right.

So our first speaker here, who is actually an expert on this field, David Wallace, who is Associate Professor of Mechanical Engineering, will talk to us about new media education. All right. Here we go. All of that being said.

WALLACE: So what I'm going to talk about is really the new media possibilities that are enabled by the use of information technology in engineering education. To provide a little background about myself, I'm the Associate Director of the MIT CAD Lab. I do my research in the area of computer-aided design. And my teaching is in the area of product design, both at the graduate and undergraduate levels.

So what am I going to do? I'm first going to provide a little bit of motivation, which is at least the angle at which I am going to be coming from in this presentation, provide a few facts of things that we've learned and that are going on, talk a little bit about some experiments that we've done in the department at MIT, and then move on to some thoughts about teaching and learning models in this new information technology enabled environment.

So to me, this represents the space in which we teach. And we're certainly used to, in universities, the notion of presence and traditional media, where traditional media may be classrooms, lectures, traditional labs, tutorials, and things like that. And of course, now into the fray we have the possibilities of using new media and incorporating tele-presence as part of our educational programs.

And so really, ideally we'd like to be able to work pretty seamlessly throughout this space. And some of the key questions are, how do we design our curricula in an informed way so that we can understand the difference in when we might want to use traditional media, or new media, or presence and tele-presence? And it's an important question, because often the tele-presence can compete or at least offer some of the similar-sounding educational opportunities at a lower cost than a residential education.

So moving on to motivation, what I think we need to be doing and are interested in doing is trying to develop and test new strategies for teaching when we have this combined media environment that can be applicable to a broad set of curricula. And we can understand when we might want to use presence or tele-presence to really have the most effective educational opportunities possible.

And what I mean by that is, first of all, we need high quality mass education. And we also need very high quality residential education. So we can clearly understand and show the value of the on-campus educational experience. So that's the motivation.

And now I just want to move on to some facts, or some I would like to pose as facts. And there's still often a lot of debate about presence, or tele-presence, or correspondents, in slightly more dated form of the same thing. And this is, in fact, the "No Significant Difference" web site. And what it has is a list of over 300 review journal and conference articles that cite no significant difference based on location, i.e. presence or tele-presence alone. It's also available in book form.

So in some sense, I think the debate about presence is really open. It's pretty clear you can do a good job with presence, and you can do a good job with tele-presence. But presence in a sense has no inherent advantage. What really matters is what learning and motivational opportunities you provide in the educational experience. So it's kind of like what do you do with the time? How well do you use the situation to provide an educational experience?

We'll also just note that there is also a growing significant difference site, where most of the papers are finding that remote education has an advantage. And one might speculate that this may be because some of those programs are fairly carefully thought out, maybe more than some of the traditional classes.

So the message really is, how things are implemented is very important no matter what you do. And it's not just whether you're doing it with information technology or with people together or remotely that's really going to be the most important factor. So you have to think carefully about what your teaching strategy is.

I think another observation is that we all generally feel that we could be doing a better job. And this is just an example from Ben Linder, who was a doctoral student in the department under Woody Flowers. But basically, the idea is that he surveyed the senior students at six major universities, asking them how much energy was in a 9-volt transistor battery. And that's a log scale on the bottom, and you see a pretty widely varying response.

[LAUGHTER]

And I mean, it's a fair debate. You could say, well, should we expect them to really be able to answer this question or not? And some people might say, yes, some people might say, no. But on the other hand, I think we can say we would like all of our graduating students to be able to look at the answer they're proposing and decide whether it's ridiculous or not. Right?

And there's all kinds of other stories about things happening. And in some ways, the students in some sense aren't that much different than they were. Maybe different experiences, but they're people just like everyone else was historically. And so we need to think about what we can do to improve the way we're teaching them.

Some other facts. Sticking our lecture notes online is something that's fairly easy to do, and it certainly has administrative advantages. But in some ways, it really has a modest educational impact. And this is probably the most common applications but it really doesn't go to the core of trying to use the technology and new opportunities to teach better.

So it's not surprising, because we need to design materials to take advantage of the media that we're using. And we need strategies and pedagogical reasons for choosing the different media. And simply stated, we design a lecture in a certain way if we're going to stand up and talk, just like I'm talking to you now.

If we were going to be doing in an electronic format with multimedia capabilities, we would design it and structure it very differently. So we shouldn't expect to be able to do just a direct port of materials from one media to the other and get big advantages.

Another area of discussion or debate is often about what is the best way to teach. And if you look at a lot of different materials on education-- and teaching engineering is one, and there's a number of books on the subject-- in some ways, there really is no singular best way to teach.

So of course, there are different types of people that like to learn using different types of formats of presentation of the material. They like to learn things in different order, like with theory first, examples first, motivation first, applications first.

And there are different use patterns. Some people like to hear a story, follow it along. Other people really just want to go in, grab the salient facts. So those are all different possibilities that are really contingent upon the people themselves in different learning styles.

There's also different types of learning, whether we're trying to do training, or whether we're trying to really teach deep theory or general concepts. And there's different types of content. And we should expect that there may be different ways to address some of these issues differently.

So it's not a question of what is the best approach. But how do we have a framework so that we can apply a lot of different tools and approaches in an effective manner? And this is one hypothesis that I certainly believe in. I'll show some experiments pertaining to in a bit that at least the new media materials certainly can be designed so that they can accommodate some of the different learning style issues.

Okay, and with that, we're now going to move on to talk a little bit about some experiments. And this is a new media teaching model that is a result of some experiments. I'll just give it to you upfront.

But the gist of it is to have students study codified material using lectures and quotes, in a sense, that are online. And then following that up with experiential activity or mentoring with someone that has experience or wisdom, if you will, that's hard to codify. And then providing the opportunity for in-practice application to really consolidate learning in a deep way.

And that particular model was based on some experiments where we first tried to sit down and design a framework that we could provide the codified materials online-- in a sense, in lieu of a lecture. And really motivating that structure around the different learning styles idea. And then once we wanted to do that, really evaluate it. How well does it compare to a regular classroom?

And if we can use these materials in lieu of a class, how can we spend time differently with students, versus the traditional lecture? So I'm just going to move on.

So we developed a general template for providing lectures. And this was for, actually, an implementation of it, but for a traditional 90-minute lecture. And to just want to note down the right side is what would be the equivalent of the lecture divided up into 94 different sub-topics.

And if you follow it through, it's more or less going to read like a story. If you go point to point, it provides you very parallel access into just about every part of the lecture. Right now, we're in a section actually related to prototyping and actually how to prime things. And in that, the content is divided up.

And it's hard for you to read. But essentially, there's a bin, which is for why do I want to know this. If there's any theory behind it, that's where the theory is provided. Procedural information, which is how you do the application-- it's really an application. An example, so the students can very quickly or conveniently, if you will, tab between different types of learning information.

And then finally, within the subject, the students-- so we're looking at how to-- so that's the procedural part of priming-- they can play a video which is annotated, so they can watch and listen. They can listen to audio and look at the pictures. Or they can go through a very traditional read and look at the pictures, just like an online textbook. And we've implemented four lectures in this way, two for product design and two related to classical control.

Once we had, in particular, the prototyping lecture implemented, we wanted to compare this to our regular classroom teaching. So we had half of the class take the web lecture and the other half have a traditional lecture on the same content. And they all had access to the same textbook. But you either had a class or you didn't.

And then they had a project, and an outside group evaluated the project. And what we could do is then track a lot of information, track how they did on the projects according to the jury, and also do focus groups to try to learn a little bit about what we're doing.

So just to compare, first of all, the jury's scoring of the prototyping project. And what this is on the left is, if we divide the class group and the web group, but we look at all of the assignments they did over the term, where they were taught in the class and the same thing. And what we can show is that they are indeed expected to perform the same.

And when we look at how they did on the assignment, in fact, there was about a 7.3% difference. So in other words, the people who only had the web-based experience did a little bit better than the people who had the class. And statistically, for these data, it's about a 1 in 100 chance of seeing that difference in the mean where they actually really were performing the same. So there's a reasonable statistical evidence to say that they performed differently.

Let's think a little bit about why. Here are some of the things we can design the online material to do. It's a classroom lecture, 90 minutes. Most lectures are either one hour or 90 minutes. Both are longer than people's attention spans. And if we look at how many [INAUDIBLE] whacks at it the students online took to cover the entire lecture, the average was about three, which corresponds to 33 minutes per session, which is very close, if you look at some of the data, to what people's attention spans are.

This little snapshot down at the bottom is essentially the number of people studying as a function of time of day. And the only slot that's quiet is between 4 AM and 8 AM. The spike is actually at lunchtime. And we also found through the focus groups that they like being able to do the studying by themselves, because they didn't feel peer pressure. They felt they could spend a lot of time where they didn't understand things, and not feel like they were not as smart as the next person in the class, basically.

And we also have data where we had these parallel delivery paths. And we can show that there were groups that were movie centric, there were groups that were text and image centric, and there were groups that would switch back and forth, probably based on the content of the section, and whether they thought a different medium might have different types of information.

So what we can't say is that, looking at these differences, one of the other nice features that when we try experiments with new media materials is that we can actually track and understand a lot of things, and learn about how to do better.

But given the notion that a lecture like that online can work, the question then becomes, how do we use the class time? And so to follow up what we did is, of course, in the next year, the entire class uses the web material. And we ask them to prepare beforehand by studying this.

But the class is divided up into two groups and taught differently in person. One group, I call it the active classroom, so it's a lecture where we're following active learning type of guidelines. But the gist of it is, we give a complete second coverage of the material, which is potentially a positive thing.

For the other group, we actually make no attempt to cover all of the material, but we just go and work with them and do some things with them that cover part of the material. Then we follow the same prototypes where they all do the same sort of project, and an external review of how they do.

And in this case, again, we shouldn't show, we would expect them to perform the same based on their other assignments. And we see a difference of about over 10%. And in this case, from a statistical standpoint, there is a chance of about 1 in 1,000 of seeing that big of a difference in mean when the two groups were actually the same. So it seems that combination works pretty well.

And again, we might think, well, why? And earlier on, I said a lot of education is about motivation, and preparation, and studying, which is important. On the left is in a survey how the students rated their preparation beforehand. And so in blue are the people that essentially thought they were coming to a lecture. And the red is the people who thought that they were going to do some things.

And of course, you can see that the people in red were much more motivated to prepare ahead. These are the actual data, because they were all studying the web lecture. And what we can see is that the 60th percentile person coming to a classroom was less prepared than the least prepared person that was going to come and do an exercise.

So there were some strong motivational differences. And in fact, the only differences in the instructions between the two groups is, one said, you're strongly encouraged to review the complete lecture before class, and the other one, we'll assume that you've covered it and we'll be working on things in our class period.

So you can use the mix of information technology and traditional approaches to provide strong motivational forces. And when we asked students about-- and then, this is in a survey, if I could have only had one of the forms of instruction. If we look at the people who had the class and the web, 60% said, well, I'd take the web only.

If we look at the people that had the experience, essentially 80% said they would take the experience only. And that was in the fact that they really wouldn't-- we didn't cover all of the information there. So if they took the experience only, they would have missed the huge part of what they really needed to learn. But they placed a very high value on that experience.

And so in some sense, one of the messages is, when we're working with the students, we should be trying to do things with them, and things that are not readily codified in other forms.

So where might this type of model apply? And so this model which we were trying to support was studying codified materials using appropriate IT-based approaches, experiential activities, and then in-practice application. And what I want to add is, each lecture took about 500 hours to produce for about an hour of lecture. So there's some overhead and preparing them.

But certainly, this seems like an appropriate form where this knowledge base is stable. There's a lot of students, or a lot of customers, If you want to think of it that way. And they may or may not be super interested in learning the material, because we can bring in very high production value.

We also actually have used this model at the Ford Design Institute for their technical employee training. And what it's probably not appropriate for is emerging knowledge, where there's a small customer base and very strong interest and motivation to learn on their own. And you don't need, essentially, some of the production value.

So now with that through, what I just want to talk and close up with are some thoughts about teaching in this new media environment. And this is based on work done by a committee with several different disciplines involved in the Department of Mechanical Engineering. How am I time-wise?

MODERATOR: You have 4 minutes.

WALLACE: Okay. Should be just about right. So when we worked on this, and what we really wanted to do is try to find a way to think about all the possibilities and how we might apply them in some different teaching strategies. And these are some methods that have really been around a long time, one which we call the Socratic method, which is independent study guided by mentoring, which is more or less what I've described.

Just-in-time, which is the notion that you're working on an application, and you learn when you need to, because you've encountered a problem. And finally, a classical scientific discovery mode of learning through observation, study and hypothesis, and experiments.

And so really, what we think needs to be done is to identify the combinations of material delivery, hard facilities, and interactions for these different types of learning models, and trying to understand what ones suit different types of materials and teaching goals.

What we envision for pretty well all of these is, one thing is that most students will have laptops. And for the Socratic, we think that we'll have these types of codified lecture modules that support different learning styles. And then we can combine them with mentoring experiences.

And these mentoring experiences may be mini lectures. They may be some form of a classical lecture. It could be personal mentoring. It could be electronic forums. It could be online quizzes with feedback. It could be a studio environment. And there's activities in all of those areas and we'd like to think about how we would use them.

And one thing I can say that I feel pretty confident about, may be a bit controversial, but at least for me, I find that 15 minutes of direct contact with two or three students is easily worth a couple hours of me standing and lecturing. So that's kind of what's driving that model.

For the just-in-time model of education, what we think is needed is essentially a marketplace of learning modules, so that students or employees are working on projects, and when they encounter problems, they can find the resources to learn the bits that they need at the time.

And the reason we think there's going to be some sort of a module marketplace is that the projects are often very different and diverse. And so it's hard to predict upfront what types of codified material you would have. And it would be, I think, too great a task for any one institute to produce all of the materials that are needed in an appropriate way.

And you essentially need highly customizable curricula. And this is one of the challenges in 2009, which is a privately designed course that we teach where different teams of students are all working on different projects. And so they run into different problems as they're working throughout the term.

And finally, the scientific discovery model. We envision-- this is a picture of the high school chemistry lab concept. But really, the classroom and the laboratory are integrated. And there's a lot of interesting issues related to the use of physical space versus virtual labs-- haptic devices, for example, to virtual simulations, or physical toolboxes that need to support this type of environment. So there's a lot of potential combinations of traditional and information technology approaches that need to be thought about.

So in summary, I think there's quite a bit to be optimistic about, as there's a lot of new teaching opportunities that are arising. And I think what we need to be doing is thinking about how we can use them effectively. The one thing I will say, though, is that things are moving very fast.

There are a number of open universities, virtual universities, companies or organizations offering courses at very low cost. And so I think now is really the time that we need to explore and develop a strategy for teaching in this environment. So that's the end of my talk.

[APPLAUSE]

MODERATOR: So questions for David? Tony, maybe we could get you to try to work on you technical project while David's answering questions.

AUDIENCE: You say the direct contact is maybe 8 times better than the lecture. How did you arrive at that metric?

WALLACE: That particular metric, that's one that's not as derived as most of the data that I put up there or as careful. And that's, to be honest, my gut feeling. And to provide some substance to that in the graduate course that I'm teaching, I now lecture-- a number of the lecturers are using this approach. And then we work on activities.

I do some traditional lectures. But just about every third week, I don't lecture. And I have 15-minute meetings with two students at a time, see what they're doing, and I work with them. And I feel that the results are better. So that's so much more qualitative statement, but that's what I base that on.

And I find that 15 minutes, to be practical, you could spend more time and it would be great. But that's the chunk of time that it becomes a manageable proposition for me as an individual doing this. So yeah.

AUDIENCE: To follow up to that, with what you describe with your technology and integrating it into the teaching, and then thinking about distance learning as an area that is growing, then back to your 8 to 1 contact. Where do you see things perhaps going? And is the distance learning going to replace residential education or what?

WALLACE: Sure. I'll offer some thoughts on that. It's, of course, speculation, just like if anyone else is answering the question. My own opinion is that first of all, I think the line between distance learning and residential learning will become more blurred. And that I think even in a residential university, some things will be done almost in a distance-learning style. Such as the very codified materials.

Examples of that-- I could look the references up. But I know of lots of cases where they've begun to offer courses on campus both internet-based and traditional. And the internet-based clearly win out, in terms of what the students want to do. So the lines will become blurred.

But I think that what the residential universities really need to do is focus on the unique part that you can't really do at a distance, so that there is still a very strong motivation for wanting to have a residential education. Because you can work essentially with your peers, and also with people that have wisdom to impart that can't be readily codified and as easily disseminated.

Now, in the distance-learning mode, you can still try to do things to help impart that. So for example, you might set up a scenario where maybe you're broadcasting a type of learning experience, where you have some students who are up and actually working and interacting, almost in a mentoring way, with the faculty. And then there's observers at remote sites. And that can help bring some of the flavor of the one-on-one experience.

But it's still, from a motivational standpoint, very different than the knowledge of the students that they're going to have to go and talk, and do these things, and be seen by others. So I don't think there'll be a complete replacement. But there may be a shift in scale and what's done in different ways.

MODERATOR: One more question, and then we'll move on to the next talk. Yeah, go ahead.

AUDIENCE: It seems easy to embrace these kinds of technologies when you're trying to present theory or topical knowledge like the knowledge part. What about the know how, like how to solve problems? Have you thought a little bit about it?

WALLACE: Yes. I think, again, actually, the point you make, I think, is a good one. And it goes back to, I think, that things that can be readily codified are very appropriate for this type of approach. Things that can't be is where the value added in the experience is. And that, in a sense, the know how that's procedural, which is, in fact, lots of times what you see in employee training programs at companies, that's just as codifiable as anything else.

The kind of wisdom of actually doing an experience with experiments, say, with someone who has a lot of depth in that field is something that you're going to have a much more difficult time packaging, if you will, in this way. And that's where I think that a big part of the residential education's value should be.

MODERATOR: I know some of you have more questions. But we should really get on, because we're behind schedule, and we may have technical difficulties. So David's going to be on the education panel coming right up. So please grill him some more.

[APPLAUSE]

PATERA: I thought David and I had agreed that there wasn't any substance in our presentations, and now he's gone.

[LAUGHTER]

PATERA: All right. [INAUDIBLE] should not worry about the sound and maybe get me the video anyhow.

MODERATOR: Okay, this, by the way, is Tony Patera. He's [INAUDIBLE] mechanical engineering. He's the deputy director for the Singapore Alliance. And he's going to talk to us today about the MIT-Singapore Alliance for Education.

PATERA: So I will start with what was going to be an audio-video, and now is going to be a video with your own personal voiceover. And not even a video at this point.

[CHUCKLING]

All right. So SMA is the Singapore-MIT Alliance. And this is a little video that we made for promotional purposes. But it also gives you a sense of the program. This was launched last year. The first class will be graduating this summer.

And there are two programs initially-- in advanced materials and in high-performance computation. And as you can see, there's a very significant distance component, that I'll be describing a little more in a moment. The program begins by the students in Singapore with about a month of various teaching activities. They then come to here at MIT.

This here is indicating what they do while in Singapore, which they recognize as MIT. The beginning of the program-- the two days is focused on the spirit of technopreneurship within MIT, with examples from the MIT faculty. Here's John Vance. As well as a continuation of the more formal activities that they began in Singapore related to coursework and project work.

There's also the usual social activities that have been arranged for the students here. There we're teaching them how to abandon healthy Southeast Asia [INAUDIBLE].

[LAUGHTER]

Here we're putting on one of the posts that went to a bottom of a lake in Arkansas.

[LAUGHTER]

AUDIENCE: Can you turn your mic on please?

PATERA: And here, the students are saying how much in particular they enjoyed my lectures during their time at MIT.

[LAUGHTER]

When they return from MIT back to Singapore, they continue on with about a month of classes at Singapore. And then begins the distance aspect of it. And a key component of the distance aspect is that you saw MIT students in the classroom there. And now, in a moment, you will be seeing the Singaporean students sitting in Singapore.

We send two streams to Singapore. One is the video stream, which is all the usual talking head business. The other stream is the data stream which is essentially uncompressed PowerPoint slides and other data aspects that therefore arrive uncorrupted on the Singapore side.

The facilities that you're seeing here are the ones at MIT. There is a mirror site at Singapore. And in fact, in some of the courses, we even beamed internally to MIT. You're seeing it now from the Singaporean-- the SMA student's perspective, where you saw the two large screens at the front of the room.

The left is the video stream, the right is the data stream. And in fact, these are the kinds of PowerPoint animations that are used in some of the lectures. And I'll be talking about that a bit subsequently, as well. This again is from the Singaporean side-- Jacob White on the left. You see him now in the inset here. And on the right was the data stream that he sends over, which you see here, as well.

And again, that's from the SMA student perspective. What was being used on the left is a document camera, which is simply a higher data integrity version of a blackboard. There is a significant asynchronous component, in which all of the various lecture materials-- digitized videos, lecture notes, slides-- are posted on websites and integrated using, effectively, a system related to the command system used at MIT.

Here you see the Singaporean students viewing Professor [INAUDIBLE]. The SQL, all right? Having seen him one slide, they now see him through the digitized videos, which we found to be a very effective way for them to be able to review material. There's also other various, by this point, the standard mechanisms for them to search through the material and to pick the various pieces that they're looking for.

In addition to the distance interactions, many of the students spend time at MIT. In particular, all the PhD students. And here, you see them commingled with MIT students on campus, and also talking about research projects, here with Professor Suresh.

Research is a significant component of the SMA program. And I'll talk about that a bit more later. Here, it's appropriate to indicate that the third program is housed primarily, but not exclusively, in mechanical engineering. You see Professor Chen and Professor Hart and their counterparts in Singapore designing the manufacturing program, which is the third of the three academic programs. There are also major players from the Sloan School in that program, as there are, in fact, in several of the programs that are currently launched.

And so I think that is about it. They're just saying how great everything is. And I think at this point I can conclude. It seems to have concluded for me. And so what I'd like to do now is make a few-- there are a few uplifting moments at the end that I spared you.

But what I'd like to do now is use that as data and make a few observations, most of which I think are quite obvious. First, it's an experiment in teaching and in research. And I think you saw that I put, "at a distance." But that's only really part of what all of this is about.

And so I'd like to first make some general remarks. The key one is that SMA is not really typical of distance efforts. The first is, there is a focus on an elite graduate cohort. The students that are admitted essentially pass the same kind of muster that we require of all our own students here at MIT, in terms of GRE's and, of course, letters of recommendation.

So again, the focus is on elite, as opposed to many other distance efforts, which are more for the unwashed masses. Both teaching and research components figure prominently. Again, many distance efforts are focused on information transfer, much more of the formal teaching, as opposed to the more labor intensive research side.

There are both distance and face-to-face elements, as I indicated. The students come here, and the faculty go there. All right? So there's a significant component of what we consider more classical modalities that are not facilitated by technology in any significant way.

There's a very high faculty-to-student ratio. Many distance education efforts are stimulated by the desire to be able to much more inexpensively educate a much larger group of students. In this case, we seem to have almost maximized the inverse of that ratio, in terms of the resources that are being devoted to these students.

There's a strong emphasis on interaction. All right? This is not a case where there's a lot of tools that are developed, put online, and they become a one-on-one activity between the student and the tool. There's a lot of interaction between the faculty and the students. There are strong links with MIT residential initiatives. Many of the programs have, for example, stimulated M-Eng degrees here on campus, based on material developed within the Singapore-MIT Alliance rubric.

There's also a very strong partnership emphasis with Singapore. There are Singaporean faculty involved, so that students, even when they're in a distance mode relative to MIT faculty, have Singaporean faculty from NUS and NTU that serve as local face-to-face faculty in the usual sense. So there's always this component of a more traditional environment.

There are very advanced facilities and technical support. This is not the kind of thing one can do in one's own garage yet. And finally, there is a quasi-antipodal geographic framework that was apparently maximized for frequent flyer miles rather than convenience by our dear provost, Bob Brown, when he set up this program.

And finally, there's, because of all that, a rather nontrivial price tag. Again, distance education sometimes is seen as the economic panacea for educating a larger and larger group of people. Certainly they should not emulate the Singapore-MIT Alliance if that is their objective.

So let me then make a few more specific observations, first related to teaching, and then just a few on research. The first on teaching is that technology is no replacement for, and can significantly detract from, scholarly, organized, complete, and effective development, by which I mean preparation and assimilation of educational material.

And to give you one example of that, this is very anecdotal, obviously. We had one course in which there was a great emphasis on more dynamic presentations, and there was not too much background material available. We had another course which had much more detailed lecture notes, but much more static and less stimulating, in some sense, lectures.

And we found that the students, although they certainly enjoyed the dynamic presentations, would much prefer that we spent our time on making sure there was always adequate codified material for them to be able to appropriate and subsequently apply the material.

This is a bit peculiar, perhaps, to the SMA program, in the sense that we're developing new graduate curricula. So existing textbooks are not readily available, and in as much, background material is more imperative developed for each course. But nevertheless, I think there is a more general message to be taken away.

Ah, I guess that is not the right button. I'd like to go back to the asterisk, which says, it can significantly detract from, but need not. We saw a number of examples that Dave Wallace showed where obviously content technology was enhancing it. I've chosen to take this more negative view, because I think there's a lot more material out there that's not exciting. And the kinds of things that Dave is doing is the exception, not the rule.

This really refers to, as I just described, content technology, and in some sense, using technology to enhance the way the content is presented. All right, there's another aspect of technology which is obviously crucial to the SMA effort, which is transmission technology. You saw there some of the live beaming sessions that are done.

We have the classroom here at MIT. And there's the MIT faculty member. There's the MIT students present. That is live, simulcast over internet to the Singapore site. Obviously having all those pieces work is absolutely critical to the success of the program. And I think the summary at this point is that advanced video conferencing technology is reasonably effective and reasonably reliable, or reasonably reliable and reasonably effective.

Issues remain. Audio and video clarity. You saw the audio and video clarity. We've actually had at least as much trouble with the audio as the video, which is a bit surprising. Bandwidth and compression are an issue. We have MIT faculty that are extremely animated. In Singapore, that comes out as kind of a throwback to Claymation after they've been sent through this low-class filter.

[LAUGHTER]

Apart from sedating our faculty, there doesn't appear to be a near-term solution to that problem. And it's something that has to be addressed. Cueing and flow is also critical. We all take for granted in an environment like this how easily it is to pick up on peripheral cues in order to know where the attention has to be focused. At a distance, a lot of that cueing goes away, partially because of clarity, partially because of camera angles, partially because of one's ability to see the remote site on a screen which tends to be rather small and at the back of the classroom.

So for all those reasons, things are not quite where they should be. But they're not bad. And the feedback is generally good, based on over 100 live beamed sessions and also on student performance.

MODERATOR: Maybe use this one.

PATERA: Yes. Is this not working?

MODERATOR: Oh, the battery's dead.

PATERA: Battery's dead. So keep the other one also?

MODERATOR: Yeah. That one's fine.

PATERA: I'm gonna get electrocuted by the end of this.

[LAUGHTER]

In terms of student performance, we did comparisons not quite as scientific as Dave's, in terms of the MIT cohort and the SMA cohort. And we found that they were statistically indistinguishable. In fact, so much so that the data looked rigged. But I think that the takeaway is not so much the data as it is the fact that the students in Singapore were able to learn the material at a rate similar and an ability similar to that of the MIT students here on campus.

Observation three is that lectures without student participation are a real waste of expensive estimating resources. And, as opposed to alternatives, such as lectures with much interaction, or recitations and tutorials. And the reason is, if we pay facility charges, we have it in a very expensive room. Technician charges, we have two or three technicians typically on each side making sure that the event transpires smoothly.

Internet 2 and ISDN access charges-- all these live simulcasts are done over I-2 or ISDN. Off-hour extra compensation, because of the fact that Singapore is 13 hours different from us. The lectures take off typically between 7:00 and 9:00 in the morning and 8:00 and 10:00 in the evening.

If you add all that up, and this is all being done for real-time sessions, and all you have is a faculty member that spends 99% of the time talking, that's not really a very good use of those resources. And in some sense, this has only made more clear what's already clear at MIT, and that's why I put the SMA in parentheses.

At MIT, there are lots of charges and costs associated with giving a lecture, as well. They're not as easily attributed to things like technicians and all the rest. But they're clearly there. And once again, you might ask, is that really a very good use of resources? And that's precisely, I think, one of the points that Dave Wallace was making in his earlier talk.

But of course, what enables all this is web archiving and integration, so that you can get the material before class, and then use the class for more interesting, interactive maneuvers. And in fact, that has been used very successfully so far at SMA. We archive the digitized lecture videos, the slides, and the lecture notes. And this provides for a very rich environment.

And when I go to Singapore, I often will circulate through the cubicles where the students sit. And in fact, you see them very often going back to these videos of the lectures to see what were the key points. And they also have access to the lecture notes, which develop things in more detail. And they seem to, that way, in some sense it acts as a replacement for the more constant and redundant interactions that would occur if they were not at a distance.

On research, I have just two very obvious observations. The first one is that most MIT faculty already engage in some form of distance research-- back to MIT while on travel, or to institutions of students of colleagues. The data for that is that it's certainly difficult to find an MIT faculty member at MIT. They must be somewhere. And wherever they are, they must be interacting with their students back at home in order to continue making progress on the research front.

[LAUGHTER]

A slightly less obvious observation is that distance research can be enhanced by really now standard technology. One-to-one video conferencing, which is just a euphemism for net meeting and variants thereof. Slightly more sophisticated-- shared collaborative applications. It's now very easy to take scribbled notes, as you would present to a student sitting across from you, put it through a scanner, have that integrated into a shared and collaborative whiteboard, so that you can annotate it, and the student at the other side-- in this case, in Singapore-- can annotate it simultaneously.

And again, that kind of a session is almost as good as what you might be able to construct with a student sitting on the other side of the desk. And finally, there are various hardware devices, tablets, and other input systems that make this shared, collaborative application usage a little easier.

And in fact, we've integrated all this together into little individual distance kits that we give to the faculty participating in the SMA program, so that they have all the access to both the software and the hardware that they need, in terms of interacting with their students.

And finally, then, by way of conclusions-- even at 12,000 miles, teaching and research remain-- well, they remain teaching and research. And that's more or less by construction. Where a lot of this program is, in some sense, not revolutionary, it's reactionary-- trying to replicate some of the very good things about MIT at a distance.

Obviously some of the things are a little more revolutionary. But we're trying to keep a balance between the two. Good transmission technology is absolutely essential. And I would say, it's not quite there, but it's almost there. As you could tell from my earlier comments, I'm not quite as enthusiastic about content technology, unless it's very carefully applied.

Finally, some regular face-to-face interactions are still crucial, in particular, for research, I think, in terms of establishing a rapport with the research student, which most of us have now done. In terms of launching the research aspect of SMA, that would be clearly impossible if there were only net interactions that were available.